Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Prognosis of child recipients of hematopoietic stem cell transplantation requiring intensive care.

Lamas A, Otheo E, Ros P et al.

Intensive Care Med 2003; 29: 91-96. [abstract]

Reviewed by Ladonna Hite, MD, Texas Children's Hospital

Review posted July 6, 2003

I. What is being studied?:

The study objective:

To determine the prognosis of patients who have had hematopoietic stem cell transplants or BMT and require critical care.

The study design:

Retrospective cohort.

The outcomes assessed:

Overall mortality at 30 and 180 days after discharge from the Pediatric ICU.

II. Are the results in the study valid?

Primary questions:

1. Was there a representative and well-defined sample of patients at a similar point in the course of the disease?

Yes. All patients were less than 18 years of age, and were admitted to the PICU between August 1991 and February 2000. Of 151 total BMT patients during this time period, 44 were admitted to the PICU in 49 admissions and these comprise the study sample. The primary reasons for admission were pulmonary, cardiovascular or neurologic deterioration, which is consistent with other studies (3,4). Patients had a variety of malignant and non-malignant diseases requiring transplantation. Patients also varied in the number of days out from BMT. The authors attempted to correct for this by separating them into groups: 0-30 days post-BMT, 31-99 days, and 100 or more days out from transplantation. Severity of illness was assessed by CCS (Clinical Classification Score). All patients were followed for 180 days post PICU discharge or until death. Several patients were readmitted to the PICU, only 5 were considered separate events (pts. had to be more than 7 days from previous admission and admitted for a different indication). Patients were excluded if they had a BMT in the past but were admitted for other reasons, such as post-op admissions or admission for a procedure.

The long study period may be an issue - treatments probably changed during that 10 year period which affected mortality. For example, as the transplant program became more experienced, their outcomes were probably better. PICU treatments evolved such as surfactant, more aggressive renal replacement therapy, and the use of barotrauma-limiting ventilation strategies all could result in lower mortality during the latter half of the study period.

2. Was follow-up sufficiently long and complete?

Survivors were defined as being alive 30 days after PICU discharge, and long-term survivors for 180 days. This will give an idea of survival from the acute process and ICU related mortality. However, it may not reflect the true outcome, prolonged survival or quality of life. Actually 180 days is probably more realistically mid-term survival, as it was defined by Rossi et al. (1) The article does not identify any patients lost to follow-up.

Secondary questions:

3. Were objective and unbiased outcome criteria used?

Yes. Survival was recorded. Quality of life, and disability data are not given. This possibly could have been evaluated with the Karnofsky score or a similar system (4).

4. Was there adjustment for important prognostic factors?

Yes, patients were evaluated based on use of mechanical ventilation, pulmonary vs. non-pulmonary disease, and number (as well as type) of other organ failures. Other important prognostic factors, such as underlying disease, type of donor, conditioning regimen, days from transplant, and number of PICU days were evaluated, but were not statistically significant.

III. What are the results?

1. How large is the likelihood of the outcome event(s) in a specified period of time?

Of the 49 admissions to the PICU, the overall mortality was 63%. Only 18 of 49 admissions, lived to PICU discharge (36.7%). When divided into subgroups, the mortality was 76.5% in the mechanically ventilated patients, and 50% in those who were not. Of note, mortality was lower in the mechanically ventilated patients who were intubated for non-pulmonary indications. This is higher than the 44% overall mortality reported by Rossi (1), but similar to other adult and pediatric studies (3, 4). Long-term survival only occurred in 6 patients (13.6%), with a median survival time of 56 months.

The authors also evaluated common risk factors and diagnosis as predictors of mortality. The only significant predictors in univariate analysis were: GVHD (grade III or IV), hemorrhagic cystitis, lung injury, respiratory, cardiac or GI failure, need for dialysis, 3 or more organ failures, male gender, and increasing age. When multivariate analysis with forward regression was performed, only male gender (OR 0.23, 95% CI 0.07-0.83), hemorrhagic cystitis (0.0, 95% CI 0.0-0.88), 3 or more organ failures (p=0.0001), and GVHD (0.1, 95% CI 0.0-0.78), were significant (odds ratios are from univariate analysis; the multivariate analysis OR's are not provided). This is consistent with previous articles, which found multi-organ failure and pulmonary failure to be predictive of pediatric BMT mortality (1,2,3).

2. How precise are the estimates of likelihood?

The odds ratio is given as the approximation of the relative risk. The confidence intervals for the four variables found to be predictive of mortality are narrow, which suggests the data is reliable. For example, with 95% certainty, the OR of death associated with aGVHD III-IV is between 0 (100% survival) and 0.78 (78% mortality). CI's weren't very precise for aGVHD III-IV, hemorrhagic cystitis, renal replacement therapy, or GI failure.

It is also unclear why their p values and CI don't match up in their univariate analysis (Table 3 in the paper) Ð several don't cross 1 but the p value was listed as NS. Likewise, some p values are significant yet the CI for the OR crosses 1. For example compare the Lung Injury p value (0.04) and CI (0.04-1.15) with the renal failure p value (NS) and the CI's (0.04-1.31) - why is one significant but not the other?

IV. Will the results help me in caring for my patients?

1. Were the study patients and their management similar to my own?

Yes. The patients seem to represent a variety of diagnoses commonly encountered in a pediatric teaching hospital with an active BMT service. The predicted survival and severity of illness prior to BMT is not given (i.e. high-risk), and morbidity of survivors is not discussed. However, the preexisting diagnosis was found not to be predictive of survival by these authors and others (1,4). The patients were likely similar, but was their management similar?

We should have some concern about applying the study results from a single institution to our patients. I would have more confidence in the results if the study reflected a broader practice base. I also wouldn't want to compare one center's experience after doing this for 25 years with another center's first 10 year experience.

2. Will the results lead directly to selecting or avoiding therapy?

No. Based on the survival of 30%, and the fact that the preexisting problems, type of cancer or induction, number of days from BMT, and infectious status are not significant indicators of mortality, it is suggested that there is not enough data to withhold care or initially identify patients who will not survive.

3. Are the results useful for reassuring or counseling patients?

The results are not reassuring but may be helpful for counseling families. It will be most helpful in discussing patients with mechanical ventilation and 3 or more organ failures, hemorrhagic cystitis, or severe GVHD. More information is needed on long-term survival, morbidity, and functional status.

References

1. Rossi R, Shemie SD, Calderwood S. Prognosis of pediatric bone marrow transplant recipients requiring mechanical ventilation. Crit Care Med 1999;27:1181-6. [abstract]; PedsCCM EB Journal Club Review
2. Todd K, Wiley F, Landaw E, et al. Survival outcome among 54 intubated pediatric bone marrow transplant patients. Crit Care Med. 1994;22(1):171-6. [abstract]
3. Hayes C, Lush RJ, Cornish JM, et al. The outcome of children requiring admission to an intensive care unit following bone marrow transplantation. Br J Haematol. 1998 Aug;102(3):666-70. [abstract]
4. Jackson SR, Tweeddale MG, Barnett MJ, et al. Admission of bone marrow transplant recipients to the intensive care unit: outcome, survival and prognostic factors. Bone Marrow Transplant. 1998 Apr;21(7):697-704. [abstract]

 

Comments

Back to the EB Journal Club Index